[U-Boot-Users] Unable to compile the U-boot-0.4.8 version for OMAP1610/1510/smdk2410

[U-Boot-Users] Unable to compile the U-boot-0.4.8 version for OMAP1610/1510/smdk2410

[Rahul Shanbhag]

Hi All,
I am not able to compile the new u-boot tar ball i.e. u-boot-0.4.8,
u-boot-0.4.7 compiles fine .
I got the following error message after
make omap1610inn_config
make command

"In file included from
/home/rahul/omap1610/0-4-8/u-boot-0.4.8/include/common.h:4
0,
                 from uboot_drv.c:26:
/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/stdarg.h:43: parse error
before
`__gnuc_va_list'
/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/stdarg.h:43: warning: data
defin
ition has no type or storage class
/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/stdarg.h:110: parse error
before
 `va_list'  "

can anybody tell me what is the problem.


Thanks
Rahul

[U-Boot-Users] Unable to compile the U-boot-0.4.8 version for OMAP1610/1510/smdk2410

[Wolfgang Denk]

Dear Rahul,

in message <002701c37923$8fc51520$d253579d@a0875497> you wrote:
>
> I am not able to compile the new u-boot tar ball i.e. u-boot-0.4.8,
> u-boot-0.4.7 compiles fine .
> I got the following error message after
> make omap1610inn_config
> make command
...
> /usr/lib/gcc-lib/i386-redhat-linux/2.96/include/stdarg.h:43: parse error
> before
> `__gnuc_va_list'
> /usr/lib/gcc-lib/i386-redhat-linux/2.96/include/stdarg.h:43: warning: data
> defin
> ition has no type or storage class
> /usr/lib/gcc-lib/i386-redhat-linux/2.96/include/stdarg.h:110: parse error
> before
>  `va_list'  "
> 
> can anybody tell me what is the problem.

This means that your cross compiler is  broken.  The  cross  compiler
must  NEVER  reference  include  files  from  the  native development
system.

Best regards,

Wolfgang Denk

-- 
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-4596-87  Fax: (+49)-8142-4596-88  Email: wd at denx.de
"He was so narrow minded he could see through  a  keyhole  with  both
eyes ..."

[U-Boot-Users] Unable to compile the U-boot-0.4.8 version for OMAP1610/1510/smdk2410

[George G. Davis]

Wolfgang Denk wrote:
> Dear Rahul,
> 
> in message <002701c37923$8fc51520$d253579d@a0875497> you wrote:
> 
>>I am not able to compile the new u-boot tar ball i.e. u-boot-0.4.8,
>>u-boot-0.4.7 compiles fine .
>>I got the following error message after
>>make omap1610inn_config
>>make command
> 
> ...
> 
>>/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/stdarg.h:43: parse error
>>before
>>`__gnuc_va_list'
>>/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/stdarg.h:43: warning: data
>>defin
>>ition has no type or storage class
>>/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/stdarg.h:110: parse error
>>before
>> `va_list'  "
>>
>>can anybody tell me what is the problem.
> 
> 
> This means that your cross compiler is  broken.  The  cross  compiler
> must  NEVER  reference  include  files  from  the  native development
> system.

Um, it's not the cross compiler that's doing it in this case. gccincdir
is not defined for drivers/sk98lin. Try the attached patch. It WFM/YMMV.

--
Regards,
George


> 
> Best regards,
> 
> Wolfgang Denk
> 
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: patch
Url: http://lists.denx.de/pipermail/u-boot/attachments/20030912/f44bba7f/attachment.txt 

[U-Boot-Users] Unable to compile the U-boot-0.4.8 version for OMAP1610/1510/smdk2410

[Wolfgang Denk]

Dear George,

in message <3F61D58D.7090308@comcast.net> you wrote:
> 
> Um, it's not the cross compiler that's doing it in this case. gccincdir
> is not defined for drivers/sk98lin. Try the attached patch. It WFM/YMMV.

Well, gccincdir is not needed for drivers/sk98lin/* -  at  least  not
with  the  compilers  used  in  the  ELDK. Also, I seriously doubt if
drivers/sk98lin is actually used by the omap1610inn board.

> --- drivers/sk98lin/Makefile.orig	Fri Sep 12 10:15:26 2003
> +++ drivers/sk98lin/Makefile	Fri Sep 12 10:11:56 2003
> @@ -25,6 +25,8 @@
>  # Makefile for the SysKonnect SK-98xx device driver.
>  #
>  
> +include $(TOPDIR)/config.mk
> +

But this patch is OK anyway. Thanks for the hint.

Best regards,

Wolfgang Denk

-- 
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-4596-87  Fax: (+49)-8142-4596-88  Email: wd at denx.de
CONSUMER NOTICE:  Because  of  the  "Uncertainty  Principle,"  It  Is
Impossible  for  the  Consumer  to  Find  Out  at  the Same Time Both
Precisely Where This Product Is and How Fast It Is Moving.

[U-Boot-Users] Unable to compile the U-boot-0.4.8 version for OMAP1610/1510/smdk2410

[George G. Davis]

Wolfgang Denk wrote:
> Dear George,
> 
> in message <3F61D58D.7090308@comcast.net> you wrote:
> 
>>Um, it's not the cross compiler that's doing it in this case. gccincdir
>>is not defined for drivers/sk98lin. Try the attached patch. It WFM/YMMV.
> 
> 
> Well, gccincdir is not needed for drivers/sk98lin/* -  at  least  not
> with  the  compilers  used  in  the  ELDK.

Hm, that's odd. FWIW, I just tested:

	ftp://ftp.leo.org/pub/eldk/2.1.0/eldk-arm-linux-x86

and it has the exact same build failure as:

	ftp://ftp.arm.linux.org.uk/pub/armlinux/toolchain/cross-2.95.3.tar.bz2

Due to:

	make[1]: Entering directory `/Projects/OSS/cvs.sourceforge.net/u-boot/u-boot/drivers/sk98lin'
	arm-linux-gcc ... -isystem /usr/lib/gcc-lib/i386-redhat-linux/2.96/include ... -c -o skge.o skge.c

W/o the patch. But I'm just being pedantic now.  : )


 > Also, I seriously doubt if
> drivers/sk98lin is actually used by the omap1610inn board.

Correct, it's not used for omap1610inn board. But it gets built
anyway and causes the build to fail due to "-isystem $(gccincdir)"
where gccincdir is not explicitly defined and is probably just
picking up a default from some builtin makerules or something?

--
Regards,
George

> 
> 
>>--- drivers/sk98lin/Makefile.orig	Fri Sep 12 10:15:26 2003
>>+++ drivers/sk98lin/Makefile	Fri Sep 12 10:11:56 2003
>>@@ -25,6 +25,8 @@
>> # Makefile for the SysKonnect SK-98xx device driver.
>> #
>> 
>>+include $(TOPDIR)/config.mk
>>+
> 
> 
> But this patch is OK anyway. Thanks for the hint.
> 
> Best regards,
> 
> Wolfgang Denk
>